Allow a plan to define multiple research outputs

[briri]

Add a new Research Outputs tab similar to the one introduced by DMPOPIDOR. The user can then define outputs based on the expected info from the RDA common metadata standard.

The narrative portion of the DMP will remain unchanged at this point in time.

Each plan can have n datasets.

A new table called research_outputs should be created that contains information derived from the RDA common metadata standard it should contain the following columns:

• *plan_id: fkey to plans table
• *output_type: one of the following enum values from the DataCite metadata schema for resourceTypeGeneral (or a subset of this list: [audiovisual, collection, data_paper, dataset, event, image, interactive_resource, model, physical_object, service, software, sound, text, workflow, other]):
• output_type_description (required if selected output_type is 'other')
• format: mime type selector (filtered by output_type selection and perhaps not visible/relevant for certain selections like ‘Event’)
• *access: one of the following enum values (default to ‘open’, options are: [open embargoed restricted closed]):
• *title: string
• abbreviation: string (added for compatibility with existing DMPOPIDOR field)
• description: text
• display_order: integer (added for compatibility with existing DMPOPIDOR field)
• is_default: boolean (added for compatibility with existing DMPOPIDOR field)
• release_date: date (perhaps default to project end date if defined)
• personal_data: boolean (leaving nil equates to ‘unknown’)
• sensitive_data: boolean (leaving nil equates to ‘unknown’)
• preservation_statement: text
• quality_statement: text
• byte_size: stored as bytes but perhaps providing the user with the ability to specify MB, GB, TB and we convert to bytes in the code
• mandatory_attribution: (sometimes the dataset owner has compulsory attribution which has to be added at the top of the dataset_citation
• coverage_start: datetime (big data can be collected for very long time but the used dataset is only for some specific time period)
• coverage_end: datetime (big data can be collected for very long time but the used dataset is only for some specific time period)
• coverage_region: string (in the past I worked with cite specific datasets and each could be specific for some country)
• mime_type_id: fkey to mime_types table

Also add a mime_types table that is populated via a Rails task/script that pulls down valid mime_types from https://www.iana.org/assignments/media-types. The table should contain the following columns:

• category: string (the type of the mime_type without the sub-type ... e.g. 'audio' if the mime_type is 'audio/mp4'. This can be used to quickly fetch appropriate mime_types from within the ResearchOuput model based on the selected output_type (e.g. 'audiovisual')
• description: string (the description of the mime_type which is often the same as the sub-type, e.g. 'mp4')
• value: string (the full mime_type, e.g. 'audio/mp4')

License, Metadata Standard, Repository and Key Resource information will be obtained via external api services and stored in the identifiers table which will allow us to define [0..m] of each if desired. (Can convert the singular research_outputs.mime_type_id to the same if needed)

[mariapraetzellis]

Responding to a first pass at the new Research Outputs tab, here's a list of things to modify/change before we share with the team:

• Fix issue with not saving work if you try and save it before filling out all required fields
• The Outputs table moves the abbreviation into the title column rather than the title. Add repository, release date, and access level to the Outputs table
• Try removing the accordion tabs to see if we can display the form in one chunk rather than parsing it out
• Hide the Attribution field in the UI
• Hide Region field in the UI
• There is an error with size/units/mime type not appearing when you select the relevant Type
• On the Research Outputs home tab, remove the text that reads "No research outputs have been defined"
• Allow user to select multiple repositories, licenses, metadata standards, etc.

Specific to the Repository selector:

• Allow user to select a subject and second tier subject, change 'institutional?' checkbox to a select-box with 'Institutional' and 'General Purpose' and also allow a free text search box
• Add pagination for results list

Security and privacy:

• Capitalize "May contain sensitive data" and "May contain PII?"

[mariapraetzellis]

Going to do licenses and metadata standards features in the next sprint. Can UAT the rest of the features once this is pushed to staging.

[mariapraetzellis]

• Once you add an output and save it, wondering if it should direct you to the table listing all datasets?